Introduction
In our first lesson, we focused on the anatomy of an argument, learning how to reconstruct its skeleton of premises and conclusions from everyday language. This set the stage for the next crucial step: not just identifying arguments, but evaluating them. Today, we move from reconstruction to judgment.
This lesson addresses a fundamental distinction in logic that serves as the primary fork in the road for all argument analysis. You will learn to distinguish between the two main categories of reasoning: deductive and inductive arguments. Each type has a different goal and, consequently, is judged by a different set of standards. Your background in computer science, where you distinguish between deterministic algorithms that guarantee a correct output and probabilistic models that predict a likely outcome, will provide a useful parallel for understanding this distinction. By the end of this lesson, you will be able to classify an argument as either deductive or inductive and apply the first level of evaluation appropriate for each.
The Great Divide: Certainty vs. Probability
At the heart of logic lies a division based on the aim of the argument. Is the argument trying to provide a 100% guarantee for its conclusion, or is it merely trying to show that the conclusion is highly probable? This question separates all arguments into two families.
- Deductive arguments aim to be ironclad. They attempt to show that if their premises are true, the conclusion must necessarily be true. There is no room for doubt.
- Inductive arguments have a more modest goal. They aim to show that if their premises are true, the conclusion is likely or probably true. They offer support, not proof.
To get a quick and clear sense of this difference in action, let's start with a short video.
Deductive vs Inductive Reasoning FLOW CHART | Valid, Sound, Strong, & Cogent
This video from the "Let's Get Logical" channel clearly illustrates the difference in the aim of deductive and inductive arguments using simple examples.
Please watch the first part of the video, from the beginning until the end of the introduction. Notice how the squirrel argument intends to be a guarantee, while the grizzly bear argument only intends to show a probability.
This fundamental split can be visualized as a flowchart, which will serve as our map for this lesson and the next few.
.svg.png)
Today, we will trace both branches of this map, learning the meaning of each term along the way.
Deductive Arguments: The Logic of Guarantees
As we've established, a deductive argument attempts to provide conclusive support for its conclusion. The standard for success is absolute: either the premises guarantee the conclusion, or they don't.
Deductive and Inductive Arguments – Ethics and Society
This text provides a formal definition of deductive arguments and introduces the crucial concept of validity.
Please read the section "Deductive Arguments" and the three equivalent definitions of validity that follow, ending just before the Socrates example.
Criterion 1: Validity
The first and most important evaluative term for deductive arguments is validity. A deductive argument is valid if its structure is such that if the premises were true, the conclusion would be guaranteed to be true.
It's essential to understand what validity is not. It is not about whether the premises are actually true. It is a purely structural property. Think of it as a well-formed piece of code that is syntactically correct and logically sound; it will run without errors, but its output depends entirely on the input you give it. A valid argument is a perfectly structured logical machine.
The following reading explains this by separating an argument's content from its logical form.
Deductive and Inductive Arguments – Ethics and Society
This section explains that validity is determined by an argument's form, not its content.
Please read from the example "All humans are mortal" through the discussion of the silly "Todd is a banana" argument. Pay close attention to how the argument's form (All A are B...) can be valid even when the statements themselves are nonsensical.
So, we can have a valid argument with false premises and a false conclusion (the "Todd is a banana" argument). The one thing we cannot have in a valid argument is all true premises and a false conclusion. The very definition of validity makes this impossible.
- Valid Argument: It is impossible for the premises to be true and the conclusion false.
- Invalid Argument: It is possible for the premises to be true and the conclusion false.
This strict, binary nature of validity—an argument is either valid or invalid, with no in-between—is analogous to the boolean logic that underpins computer architecture. An argument's form either works or it doesn't.
Criterion 2: Soundness
Of course, we want more from an argument than just a good structure. We also want it to be based in reality. This brings us to the second evaluative criterion: soundness.
An argument is sound if it meets two conditions:
- It is valid.
- All of its premises are actually true.
A sound argument is the gold standard of deductive reasoning. Because it is valid, its conclusion is guaranteed by its premises. Because its premises are true, its conclusion is also guaranteed to be true. The classic example you read about is both valid and sound:
P1: All humans are mortal. (True)
P2: Socrates is a human. (True)
C: Therefore, Socrates is mortal. (Must be true)
The "Todd is a banana" argument, while valid, is unsound because its premises are false. An invalid argument is always unsound.
Inductive Arguments: The Logic of Probability
Most arguments we encounter in daily life, science, and politics are not deductive. They are inductive. They don't aim for certainty but for probability. Their conclusions are meant to be likely, reasonable, or plausible, given the evidence.
Criterion 1: Strength
We do not use "valid" or "invalid" to evaluate inductive arguments. Instead, we talk about their strength. An inductive argument is strong if its premises provide good reason to believe its conclusion is probable. It is weak if they do not.
Unlike validity, strength is not an all-or-nothing affair. It exists on a spectrum. An argument can be very strong, moderately strong, or very weak.
The crucial difference between inductive strength and deductive validity lies in how they react to new information. The validity of a deductive argument is fixed by its form and cannot be changed by new evidence. The strength of an inductive argument, however, can change dramatically as new premises are added.
Deductive and Inductive Arguments – Ethics and Society
This text introduces inductive arguments, their evaluation in terms of strength, and how they differ from deductive arguments when new evidence is introduced.
Please read the sections "Inductive Arguments," the definition of strength and weakness, and the subsequent section on how new information affects inductive arguments. The airplane crash example is an excellent illustration of this key concept.
This dynamic nature of inductive reasoning—updating the likelihood of a conclusion based on new data—is the foundation of many fields, from scientific inquiry to the probabilistic models used in machine learning.
Criterion 2: Cogency
The inductive equivalent of soundness is cogency. An argument is cogent if it meets two conditions:
- It is strong.
- All of its premises are actually true.
A cogent argument gives us a good reason to believe its conclusion is true. For example:
P1: Almost every student in this CSE program has strong quantitative skills. (True)
P2: Anjali is a student in this CSE program. (True)
C: Therefore, Anjali probably has strong quantitative skills.
This is a strong argument with true premises, so it is cogent. The conclusion isn't guaranteed—Anjali could be the rare exception—but it is very likely. If, however, P1 were "Almost every student in this history program has strong quantitative skills," the argument would be weak and therefore uncogent, even if P2 were still true.
How to Tell Them Apart
A common but mistaken shortcut is to say that deduction reasons from general to specific, and induction from specific to general. While this pattern often holds, it is not a reliable definition. The following reading explains why and provides better ways to distinguish the two.
Deductive and Inductive Arguments
This article corrects a common misconception about deduction and induction and provides a more reliable method for distinguishing them.
First, read the section "Specificity and Generality" which debunks the "general to specific" myth by showing counter-examples. Then, read the section "How to Distinguish" which offers a much better test: can additional evidence change the certainty of the conclusion?
The most reliable test is to ask: Does the argument aim for certainty or probability? Or, as the reading suggests, ask yourself: If you were to add another premise, could it possibly make the conclusion more or less likely?
- If new information cannot change the logical link between premises and conclusion, it's a deductive argument.
- If new information can strengthen or weaken the link, it's an inductive argument.
For a final review that ties all these concepts—validity, soundness, strength, and cogency—together, please watch this comprehensive video.
1.4 Validity, Truth, Soundness, Strength, Cogency
This lecture by Mark Thorsby covers all the key evaluative terms for both deductive and inductive arguments.
Watch the video in its entirety. It will serve as a great summary of everything we've discussed. Deductive vs. Inductive Validity and Soundness Strength and Cogency
Conclusion
Today, you have navigated the primary division in logical reasoning. You've learned that arguments are evaluated based on their initial aim: deduction for certainty, induction for probability.
Here are the key takeaways:
- Deductive arguments aim to provide conclusions that are guaranteed by their premises. They are evaluated as valid (perfect structure) or invalid. A valid argument with all true premises is called sound.
- Inductive arguments aim to provide conclusions that are made probable by their premises. They are evaluated on a spectrum from strong to weak. A strong argument with all true premises is called cogent.
- The key difference in practice is that the strength of an inductive argument can be altered by new evidence, while the validity of a deductive argument cannot.
You now possess the complete, high-level toolkit for classifying and evaluating any argument you encounter. In our next lesson, we will zoom in on the deductive branch of our flowchart. You will learn to more formally assess the validity and soundness of deductive arguments by recognizing specific, common argument forms.